1. Field of the Invention
The present invention relates to a method and apparatus for extracting an optical clock signal, and more particularly, to a method and apparatus for extracting an optical clock signal using a bi-directional amplifier and a Fabry-Perot filter.
2. Description of the Related Art
With the increase of transmission speed in optical communication and the development of transmitters converting data into optical signals, the improvement is also needed in a signal processing rate of a receiver, which receives an optical signal and converts it back to the original data. To satisfy the request, methods and apparatuses for extracting an optical clock signal have been devised.
To extract an optical clock signal a method using a self-pulsating laser diode, a method using an optical loop mirror, a method using an optical tank circuit, etc. have been devised. However, it is still difficult to manufacture an optical element for extracting a desired clock signal and an optical system is still unstable.
To overcome these problems, a method of recovering a clock signal using a frequency component existing in an optical spectrum has been suggested. In other words, two adjacent frequency components corresponding to the data transmission rate of a received optical signal are extracted and a beat frequency component, corresponding to the difference between the two spectral lines, is generated, whereby a clock signal is recovered.
In the above-described conventional method, two frequency components are selected from an optical spectrum and adjusted to have the same intensity. Thereafter, a beat frequency component is obtained from the two frequency components, thereby obtaining a clock signal for an optical signal. To select two frequency components and adjust them to have the same intensity, a conventional method illustrated in FIG. 1 is used.
FIG. 1 illustrates a conventional circuit for extracting an optical clock signal using a tunable band-pass filter 120. Referring to FIG. 1, in order to make first and second frequency components or second and third frequency components have the same intensity in an input frequency spectrum 110, the intensity of the second frequency component should be decreased.
For example an input optical signal with the input frequency spectrum 110 is passed through the tunable band-pass filter 120. The tunable band-pass filter 120 adjusts amplitudes of frequency components, thereby making the first and second frequency components or the second and third frequency components have the same intensity. In detail, the tunable band-pass filter 120 puts the first or third frequency component at a point P1 giving the least attenuation and puts the second frequency component at a point P2 giving the most attenuation to make the first and second frequency components or the second and third frequency components have the same intensity. Reference numeral 130 denotes the characteristic of the tunable band-pass filter 120.
Here, a difference between the intensity of the first frequency component and the intensity of the second frequency component or between the intensity of the second frequency component and the intensity of the third frequency component must be similar to a difference between the attenuation at the point P1 and the attenuation at the point P2 in the tunable band-pass filter 120 to make the first and second frequency components or the second and third frequency components have the same intensity to an acceptable accuracy. When a difference between the intensity difference and the attenuation difference is great, the method illustrated in FIG. 1 is not efficient. In other words, the tunable band-pass filter 120 suitable to the characteristics of the optical spectrum of the input optical signal needs to be used or the tunable band-pass filter 120 needs to be specially manufactured to be suitable to the characteristics of the optical spectrum of the input optical signal. Reference numeral 140 denotes an optical spectrum of the optical signal that has passed through the tunable band-pass filter 120.
Moreover, in the method illustrated in FIG. 1, an extracted clock signal is greatly influenced by the pattern of an input optical signal. In other words, when data of the input optical signal is continuously “0” or “1”, a clock signal component may disappear.